The face of the oldest known Homo sapiens, reconstructed from the Jebel Irhoud fossils

In 2017, two papers in Nature pushed the origin of our species back by one hundred thousand years at a stroke and named a Moroccan site, Jebel IrhoudJebel IrhoudA prehistoric site in Morocco, west of Marrakesh, which yielded the oldest known Homo sapiensHomo sapiensThe present-day human species, which emerged in Africa around 300,000 years ago, the only surviving human lineage after the extinction of Neanderthals and Denisovans.→ fossils, dated to about 300,000 years ago.→, as the oldest known cradle of Homo sapiens. One question science was reluctant to answer remained: what did this human of nearly 300,000 years ago actually look like? In September 2026, an international team led by Brazilian designer and facial reconstruction specialist Cícero Moraes offered a reply. Published in the journal Heritage (MDPI), their study presents a facial approximationfacial approximationThe reconstruction of an individual's face from the skull, based on soft-tissue data; it aims at a plausible appearance, not an exact portrait.→ of the Jebel Irhoud composite skull, together with a repeated warning: this face is an anatomically coherent hypothesis, not a certified portrait.¹

The researchers produced two complementary renderings. The first is a greyscale head, without hair or facial hair, designed to show as closely as possible the relationship between the soft tissue and the bone that carries it. The second is a coloured version, more familiar to the public, with skin tone, hair and a beard. This second face is explicitly artistic: skin pigmentation, lighting and hair rest on no evidence drawn from the fossil, but on conventions informed by the study of ancient and modern populations. The authors sum up their approach with a cautious phrase: to offer a "morphologically plausible" representation, strictly bounded by the anatomy of the skull, without claiming to reconstruct the individual's real appearance.¹

Jebel Irhoud, the Moroccan cradle of our species

The site's history is one of a long misunderstanding. In the early 1960s, a miner extracting barite at Jebel Irhoud, about a hundred kilometres west of Marrakesh, uncovered a skull with strikingly human features. Named Irhoud 1, the fossil was first taken for an African variant of Neanderthal, around 40,000 years old. Successive datings then pushed its age back, to between 100,000 and 200,000 years in 1991, then towards 160,000 years in 2007. The deposit, originally a cave whose covering rock had been stripped away by mining, resisted any comfortable classification.⁵

The turning point came in 2017. In a first Nature paper, Jean-Jacques Hublin and colleagues from the Max Planck Institute for Evolutionary Anthropology described new fossils and assigned the whole assemblage to Homo sapiens, within a "pan-African" origin for our species. In a second paper, Daniel Richter, Shannon McPherron and their team dated the site, using thermoluminescence applied to heated flints, to about 300,000 years.³⁴ Until then, the oldest securely dated remains of Homo sapiens came from Omo Kibish in Ethiopia, around 195,000 years. Jebel Irhoud thus pushed the birth of our lineage back by roughly one hundred thousand years and shifted the spotlight from East AfricaAfricaThe cradle of humankind: the continent where the first hominins appeared, then Homo sapiens around 300,000 years ago, before the expansion to the rest of the world.→ to the continent as a whole. The message of the two papers was that our species did not emerge in a single garden of Eden, but took shape gradually across a vast African territory, in populations that were already exchanging genes and ideas.

Excavations carried out from 2004 by the Max Planck team and Morocco's National Institute of Archaeological Sciences and Heritage raised the number of human remains from six to twenty-two, spread across several individuals. It is this series, recovered between 1961 and the campaigns of the 2000s, that now supplies the material for the reconstructed composite skull. The find also revealed a Middle PleistocenePleistoceneThe geological epoch of the great ice ages (c. 2.6 Ma–11,700 BP), spanning most of human prehistory.→Middle PleistoceneGeological subperiod (about 770,000 to 126,000 years ago) marked by growing behavioural complexity among hominins.→ stone industry, marked by the Levallois techniqueLevallois techniqueA Middle Palaeolithic flint-knapping method: a core is prepared so a flake of predetermined shape can be struck off in one blow; a hallmark of Neanderthal skill.→, which signals the beginning of the African Middle Stone AgeMiddle Stone AgeA long phase of African prehistory (c. 300,000 to 40,000 years ago), the African equivalent of the Middle Palaeolithic, marked by the rise of symbolic behaviour.→: the site is not only the oldest anatomical cradle of our species, it also bears witness to early changes in behaviour.⁵

Reconstructing a face 300,000 years old

The method used by the 2026 team drew only on free software and a transparent processing chain, a point the authors stress. The work does not rest on direct access to the original fossils, but on data the Max Planck Institute had made public in 2017, including a video turning around the digital reconstruction of the Jebel Irhoud skull. The researchers converted this video into a sequence of images, then rebuilt a three-dimensional mesh of the skull by photogrammetryphotogrammetryTechnique reconstructing a 3D relief model of an object from many photographs taken at varying angles.→, using a module integrated into the Blender software. Lacking a built-in scale marker, they calibrated the dimensions against a ruler present in the original documents, then checked the model's fidelity by comparing it with published measurements and with another mesh, obtained independently by photogrammetry and shared online.¹

Then comes the delicate stage: moving from bone to face. The team used the CT scan (medical scanner) of a modern virtual donor as an anatomical base, then deformed it to match the morphology of the Jebel Irhoud skull, in a technique known as constrained anatomical deformation. The idea is to propagate the adjustment of the skeleton to the soft tissue, so that the final shape is dictated above all by the underlying bone. The soft-tissue thicknesses applied come from datasets established on populations of African ancestry, judged the most relevant geographically and evolutionarily. A nasal projection template, drawn from measurements on living individuals, completes the estimate of the nose's relief.¹

This choice of references immediately raises the main limitation of the work, which the authors do not try to hide. There is no direct data on the soft tissue of Middle Pleistocene hominins: no skin, no preserved muscle. The only references available are those of modern populations, used as a necessary approximation rather than as a biological equivalence. The authors recall that features such as the robustness of the brow ridge or the unusual proportions of the face may have gone with a musculature and a skin different from ours. Anatomical deformation, by bounding the result with the geometry of the skull, limits this bias without eliminating it.

A modern brain in an archaicArchaicRefers to an ancient, now-extinct human population or form (Neanderthals, Denisovans, ghost lineages), as opposed to anatomically modern humans.→ skull

Beyond the face, the method produced a figure for the brain. By segmenting the internal cavity of the skull, the team measured an endocranial volume of 1,364 millilitres. This figure falls in the middle of the range established by earlier studies carried out directly on the fossil, between 1,305 and 1,480 millilitres, and is close to the 1,375 millilitre measurement obtained in 2018 by CT scanning. Converting this endocranialendocastA cast of the internal cavity of the skull, whose volume allows the brain size of a fossil individual to be estimated.→ volume into brain volume, through a reduction of about 9.8 percent, gives close to 1,230 cubic centimetres, a value very near the average for modern adult males, estimated at around 1,234 cubic centimetres. In other words, the brain of this archaic Homo sapiens had already reached a size comparable to our own.¹⁶

But size is not shape. That is the whole lesson of Jebel Irhoud, already underlined in 2017: while the brain volume is modern, the braincase itself remains archaic, low and elongated towards the back, where the skull of modern humans is more globular. This dissociation suggests that brain size was reached early in the lineage, but that its shape, and perhaps its internal organisation, went on evolving afterwards. The Jebel Irhoud skull thus illustrates a moment of mosaic: features that are already ours live alongside traits inherited from a more distant past.

The contrast is sharper still on the face. The measurements taken on the model show a remarkably broad face. The distance between the frontomalar points, which frame the upper part of the face, reaches about 118 millimetres, against an average of around 97 millimetres in modern populations, nearly five standard deviations above the expected value. The jaw too is wider than the modern average, the chin and brow ridge trace a relief different from ours, and the reconstructed cranial circumference, about 60 centimetres, clearly exceeds the modern male average of 56.4 centimetres. The picture that emerges is coherent: a brain of modern size housed in a skull, and behind a face, that are still deeply archaic.¹

An approximation, not a portrait

This is the thread running through the study, and it deserves to be taken seriously. At every stage, the authors recall that their reconstruction is an approximation conditioned by the references adopted, not a photograph of the past. Three main reasons impose this caution. First, the complete absence of fossil soft tissue, which forces them to borrow skin and flesh thicknesses from modern populations. Next, the composite nature of the skull itself. Finally, the frankly speculative share of the colour and hair elements, acknowledged as artistic.¹

The reconstructed skull is indeed not that of a single individual. It is an assembly, structured around Irhoud 1, the fossil found in 1961, which supplies the braincase and the upper part of the face. A mandible from another individual, together with dental elements from other fossils of the same layer (Irhoud 10, 11, 21 and 22), fill the missing areas. All belong to the same population, the same archaeological level, and show an anatomical compatibility that gives the model greater robustness than a purely theoretical projection. The authors note, however, that the most representative part of the face comes from Irhoud 1, and that the dentition, a proportionally small region, weighs little on the final result.

As for the dark skin, hair and beard of the coloured version, they are nothing more than a plausible proposal for a North African individual of about 300,000 years ago, resting on anthropological conventions and not on the fossil. The authors in fact produced the greyscale version precisely to prevent these aesthetic choices from drawing attention away from the anatomy. The dual rendering reflects a deliberate tension between scientific rigour, embodied by the grey head, and the need to speak to the public, served by the coloured face. For an informed reader, the lesson is clear: to look at this face is to look at a working hypothesis, not the portrait of a person.

An open 3D model and what it signals

The other contribution of the study is methodological, and it may matter more than the face itself. The whole processing chain rests on free software, and the files produced are downloadable. The team put the simplified mesh of the bust online so that other researchers can carry out their own analyses of anatomical coherence, and compared it with a skull model digitised independently by a third party, with which it aligns closely. This transparency, rare in a field where reconstructions often remain opaque, allows criticism and reproduction, two fundamental requirements of the scientific approach.¹

The technique was not, moreover, invented for this particular case. The authors recall that it rests on long work, carried out since the early 2010s and tested in varied contexts, from approximations of historical figures to forensic cases, by way of extreme anatomical conversions between primate species intended to test the robustness of the method. This accumulation suggests that the same approach could be applied to other African Pleistocene fossils, often fragmentary and dispersed among museum collections, for which direct access to the original pieces is difficult or impossible. Where a published scan or even a documented video exists, a reproducible facial approximation now becomes conceivable, at modest cost and open to scrutiny by any laboratory.

One should be careful not to see this as a resurrection. What Jebel Irhoud, 2026 version, delivers is not the true face of the first Homo sapiens, but a coherent proposal, debatable and debated, that puts features on a statistic and a shape on a skull. Its main merit is perhaps to make tangible the paradox discovered in 2017: about 300,000 years ago there lived, in what is now Morocco, a being whose brain had already reached almost our volume, but whose broad, robust face reminds us how long, how branching, and how far from complete the history of our species still was when the first Homo sapiens opened its eyes.